[christofer]

I went with the spacer method. Over the last few days I have been in the machine shop at the school making parts.

I first used some 600 grit sandpaper to polish the bearing journals on the spindle for a sliding fit. The bearings slide on with oil and bind without oil. I imagine I am within a few tenths of a perfect fit!

I cut and bored the spacer next. I also cut the short spacer for the top of the spindle between the nuts and the ball bearings. The spacers are not loose. They are no more than 0.002" too big on the inside and the inside bore is concentric to within 0.001" of the outside surface. The flat edges of the spacers are parallel to a few tenths over their surfaces. The short spacer was made in one operation without removing the material from the chuck.

Can someone suggest a spline for me to use? I can't seem to find any ready made splines. The spline on the spindle is 6 raised parts that are each 5mm wide. The OD of the spline is 25mm.

I really think I may clamp a pulley to the outside of the spline. Cost is key here. More than $30 for the spline nut is probably out of the question when I can make a clamp for free.

I love pictures of other people's builds, so here are some of mine.

[handlewanker]

Hi Chris, you have been busy.....it's starting to look good.

I once cut a spline in the lathe using the end of the male spline to "mark" the bore so that I could plunge each spline by racking with a boring bar on it's side in the lathe toolpost.....the tool bit was shaped like a grooving tool with a square end.

I first banged the pulley down on the spline shaft to get a decent indentation, then cut the grooves as marked .005" undersize on the width.

Once again I banged the pulley down on the splined shaft, but a bit harder this time and got a good indentation that I was able to replunge the groves with a size for size tool in the boring bar and finally fit by applying a file to the sides of the grooves and banging the pulley onto the splines until it was a light press fit.

The pulley was held on by a grub screw in a groove ground on the spline shaft.

While plunging the lathe chuck was held stationary by a block of wood and 2 wedges under it.

Later we made a special split band clamp that went round the lathe chuck and was clamped to the lathe bed and that ensured the spindle was rigid......we also marked a whole seies of lines (24) on the OD of the chuck backplate to act as an indexing mechanism.

Splines cut this way are very easy to do provided you can index them accurately.

The first one I ever did was with a tool in a boring bar in the Bridgeport mill spindle, plunging up and down with the side handle and the job held in a 3 jaw chuck on a rotary table, but somewhere along the line the spindle moved and we had a scrapper.....the second one I did, I made a double band clamp to go round the spindle end and clamped to the quill end housing.
Ian.

[Coxsteve]

Regarding the use of belt drives for various applications as many will read this topic that are toying with different designs and machine builds.
With many belt designs there are set tensions to be applied to designs as listed by the belt suppliers,These vary depending on center to center belt materials etc.

Many manufacturers have dedicated belt tension deflection gauge tools usually spring loaded and to suit the particular application,This tends to make these tools specific to the design at hand and therefore becomes somewhat hit and miss for custom builds.
These require some research when designing and as handlewanker pointed out earlier when talking about bearing preload and overheating of a tool grinder he manufactured for a boss in the past usually some trial and error adjustments.

Once you have found the ideal belt tension for your design there is an easy method to utilize to keep the adjustment where you want it.
I work on among other things Ducati race bikes and Ducati measure belt tension by measuring the frequency.
The factory special tool is quite expensive and so many have come up with an effective and cheap method for measuring frequency on belt drives to gain repeatable settings on belt tension.
This involves a laptop with sound card(Basically all of them) a microphone plugged into the microphone jack and some easily available free opensource software to measure the microphone input,There you have it easily checked belt tension and repeatable which is most important for consistency.

A google search on that topic will soon divulge the required details as like in most cases it is simply knowing what question to ask.

With so many designs using belt drives these days I thought that having already solved that issue elsewhere the method needs a wider audience,And yes it works with many Car engines as well when it comes to home maintenance and is less haphazard than most methods used (With a pinch of salt) by many,even so called professionals.

[christofer]

I bought two of this ballscrew on ebay. I will machine the ends to accept angular contact bearings and I will machine the blocks to house those bearings. These screws will be used for the Y and Z. I intend to leave the floating end of the Z axis screw unsupported for more travel. I suppose I will support the floating end of the Y screw though I am not sure if I really need to. I will have some machining to do on these. For $85 each it was hard to pass on these screws. They have a lead of 0.625" so I will use a 1:2.5 or 1:3 belt reduction. I am a bit scared of the backlash the belt might introduce but I have been told by people here that it will not be a problem if I get the tension right. I like the acoustical method. I have an oscilloscope and a spectrum analyzer, though I am not sure the spectrum analyzer goes low enough (it is for the low microwave range). The PC sound card method may work as well.

The design of a motor controller for the spindle is well underway. I will have an LCD display that shows real motor RPM and set motor RPM. There will be a 1000 count encoder mounted on the spindle motor or belt. This encoder provides feedback to a microcontroller on the control board. The driver uses a synchronous buck converter arrangement to smoothly drive the motor. This seems to be how high quality treadmill motor drivers work. So far the driver will fit in a 4.5 x 6 x 4 inch box and can be mounted right on the head.

The spindle housing has been milled square and I have started boring the clearance hole.

I contacted a local steel shop and I was told they can get everything except the 12 x 8 tube that my design calls for. I will likely change my design to use two side by side pieces of 8 x 8 square tube.

How much torque is recommended for a design of this size? Will 1000 in oz motors work well? I found this motor that I intend to use.

I am still in need of an X axis screw with at least 24 inches of travel.

[jsheerin]

I work on among other things Ducati race bikes and Ducati measure belt tension by measuring the frequency.
The factory special tool is quite expensive and so many have come up with an effective and cheap method for measuring frequency on belt drives to gain repeatable settings on belt tension.
This involves a laptop with sound card(Basically all of them) a microphone plugged into the microphone jack and some easily available free opensource software to measure the microphone input,There you have it easily checked belt tension and repeatable which is most important for consistency.

I do the same thing on my Ducati. My only addition is that you don't even need a laptop / mic / software. You can get a free guitar tuner app for a smart phone and it will work just as well as long as it displays the actual frequency it's recording.

[christofer]

Well... here is my first attempt. I suspect it will need to be remade as I cut the 72mm recess to 72.05mm. I also suspect the bung method does not work very well as the hole in the chuck had about 1 thou of runout.

I would really like to make the piece out of round stock as I could turn the outside down and then indicate it after I flip the piece. This way I would be sure everything is concentric and aligned.

I am currently trying to figure out how to attach a plate to the side of a round housing. I may end up bolting a plate on to the front of the spindle and mounting that to the machine.

Anyway, I am having a great time practicing my machining skills on a big lathe.

[RICHARD ZASTROW]

Can you bore that 72mm oversize then install a shrink fit sleeve? If so, leave a small amount of finish stock for a corrected bore size and location. Just a thought.

Dick Z

[handlewanker]

Hi Chris, I would not worry if the chuck bore is .001" out, if you get the second side true to within .001" you've done well......that's .001" runout over a distance of 6".......you'll never be able to detect it on the job.

Most people who do not practice engineering as a way of life would find it very hard to measure to within .001" on a bore size of 72mm........ you develop a feel for the inside micrometer or telescopic gauge which you use to measure the bore only after several years of practice.

This is not to say that any old size within the size range is OK, it means no two people can measure either an OD or an ID to the same dimension......that has been proved.

BTW, what clearance are you giving the AC bearings in their housing......they have to be a slide fit, I.E. some clearance, and the rule of thumb for clearance fits is .001" per inch of diam, tighter than that and it becomes a press fit.

Press fits are also made at .001" per inch of diam to give a press fit without bursting or distorting the inner race.

So for a 72mm diam bearing you need +.0025" clearance in the AC bearing housing bore to allow a sliding fit....the bearing outer shells are retained by the front cover, and you need at least -.002" on the depth of the AC bearing bore to ensure the cover nips the AC bearing outer race when the retainer screws are tightened.....if you have too much nip here you'll distort the cover face.

On the topic of a round housing, the retainer aspect is not too much of a problem, you can make a retainer with two split rings machined onto the Z axis face, like they do when a router round body motor drive is mounted.

You'll have the same "flip over" set-up problem with a round housing as you would with a square one.....chuck jaws do not allign round work perfectly due to bell mouthing, and you cannot set the back outside diam true anyway as it's within the chuck jaws.

It's looking good so far...too good to give up and start again.

If the AC bearing outer races are too slack in their housing, you could "recover" the job by applying Locktite to the outer races, but only on the very final assembly, to take up the slack, but this is a fix me up for a one off job.
Ian.

[christofer]

Perhaps I didn't do too bad then? I bored the 72mm AC recess to 72.05mm so I suppose I am at just about 0.002 inches of clearance. I expected to have to get this much tighter for some reason.

I ordered the round stock for a second spindle so I may go ahead and make one anyway. I may just end up with two spindles. Its really not much of a problem to have a second try. I rather like using the large lathe.

I hadn't thought about the clamp method. Coincidentally I just ordered an R8 slitting saw holder and two slitting saws.


I now believe I will be using servo motors rather than steppers. I have these in mind: High-Torque Stepper Motor, Stepper Motor, Driver, Stepper Motor kit, DC Servo Motor, DC Servo Motor kit, Stepper Motor Power Supply, CNC Router, Spindle, and other Components. Automation Technology Inc.

And these drivers: http://www.automationtechnologiesinc...r-driver-g320x

[handlewanker]

Hi Chris, if you had the AC bearings a tight fit in their bore, you would have a difficult job getting them out, but a pair of small recesses on the back of the bearing face seat would allow you to use a hook with a drop or slide hammer to pull them out if needed.

Most designs have the AC bore back recess, bored through slightly bigger than the inner diam of the outer race, so that you can push them out with a steel rod/punch from the top end of the housing.

By round stock for the second spindle, I assumed you meant for the housing as opposed to the actual spindle which you bought in.

Making a spindle per se would take a big degree of expertise due to the end having the chuck/collet mounting needing to be hardened and ground and running very concentric, and if you wanted splines for the drive....more work again.

The clamp method is just two thick plates, steel or alluminium, bolted or welded to the Z axis face and both bored through to suit the housing diam....the plates being split on the halfway line and fitted with spacer shims between the faces and also clamp screws either side of the bore before final boring....the shims allows you to tighten up the clamp to the housing when the shims are removed.

The square housing needs to have two sets of tapped holes, top and bottom to bolt it directly to the face of the Z axis.

One face of the housing therefore needs to be faced flat and true to the spindle axis for attaching to the Z axis face.....a cross key in the face of the housing and Z axis face would ensure accuracy for removal and replacement....easier than putting in dowels.
Ian.

[packrat]

@Handlewanker - you guy's sure do things different downunder. Can you give a reference to the .001/inch for press and slip fits? In my 49 years as a die maker/machinist I've never heard those figures.

[RICHARD ZASTROW]

If at all possible, comply with the bearing manufacturers spec's for the bore clearance. As Ian said, the bearing retainer will "pinch" on the outer ring's face to secure the bearing.

The removal grooves are also excellent advice.

I do not like the Locktite idea, seems like using Bondo body filler.LOL I tend to use sealed spindle bearings. The heat required to remove most Locktite and bearing "fixers" melt the seal and the lube on some rather pricey bearings.

Dick Z

add: My GMN catalog engineering section tolerance for housing bore for P4/ABEC 5, 50 to 80 mm OD bearing, 0 to +.00043". The surface finish required is better than a 32 micro-inch. That's why our bearing bores are finished on a jig grinder.

For a DIY'r I,d suggest subbing it to a grind shop.

[handlewanker]

Hi Rich, I quite agree with the Loctite warning and heating, but the bearing housing in question might have been bored a wee bit "loose".

The only real option apart from reboring and sleeving (and possible having a loose OD or ID on the sleeve too) would be to add a "gap filler" such as loctiting the bearings into the bore, mostly to ensure a "good fit", but as the AC bearings are being retained by the cover, the Loctite may be considered as a fix me upper for a loose bore.......a couple of thou oversize would not register as a problem once the Loctite is evenly distributed round the races.

Wouldn't work in industry for a customer, but for a DIY'er who wanted to get the show on the road and improve by experience for the next time, Loctite can be the saver of the day.

Packrat, on the topic of fits and clearances, the "rule of thumb" in my day was if you wanted a light press fit.....the bore or shaft size was size for size.......a tighter fit would be .001" per inch of diam for the shaft to fit bearings to, same for housings.

Clearance for running fits, EG where you had to assemble a part to a shaft like a sleeve etc, the bore clearance for the part was also .001" per inch of shaft diam.

These sizes were always used for bronze bushes, but in the case of a bush you added the amount of press fit to the bore clearance to cater for the closing down of the bore after pressing in.

In one particular instance where I worked we had to fit steel tires to hubs on steel wheels for our railway locos....the tires being 36" in diam.

The tires were shrink fitted to the wheels by heating them with large kerosene blow lamps to expand the tire....the diam being 36", the calculated shrink was -.036".

When the old worn tires were pressed off for renewal it took 150 tons of force using an old horizontal hydraulic power press.

You had to wear hearing protection to protect your eardrums from the shock wave when the tire finally broke loose....at 150 tons....it was like a canon shot, and after several wheels you had to have a rest from the effects of the concussion on your chest cavity.....created a feeling of nausea.

The rule of thumb can be understood when you try to push a 1" diam shaft into a 1" diam hole...it won't go, but if you have a bore .001" bigger it will.

You always lubricate the bore before pressing in bearings or parts etc, or you'll get metal to metal welding (fretting) especially when both of the metals are soft like mild steel etc.

The bearing manufacturers supply charts recomending sizes for various fits, and if available should be followed.
Ian.

[RICHARD ZASTROW]

Unfortunately, we deal with "industry". A shrink fit sleeve will usually stay put. If you are worried about that , fix it in position with a "dutchman" pin or screw. You can also build up the bore ID with brush plating.

We do not use "mild steel", it's 4140HT. Reasonably machinable but quite rigid & stable. Have changed some to fine grained cast iron though.

Obviously I've been at this too long. I think everything should be be built to my rather tight tolerances, I design machine tools, not toasters.LOL Yes, there are DIY short cuts, I just can't do that. (stubborn old fart)

Dick Z

[wizard]

I come at this from a different angle then some as I work production maintenance, but people need to realize that Loctite can be a life saver. It might not be ideal for a new machine build as they should be properly toleranced to begin with, but wise use of Loctite can keep people working and often lasts as long as a proper rebuild / build.

As to this spindle consider a trip to the bearing manufactures manual for proper advice on tolerances for bores. Information from the horses mouth is hard to argue with. Generally though you need to be careful not to have a too tight fit. You don't want assembly of you spindle closing up tolerance and deforming the outer race.

If you do end up finding that the bore is oversized more than recommended or that you can accept and Loctite is personally offensive, hard chrome plating might be an economical approach to resizing the bore.

[wizard]

Unfortunately, we deal with "industry". A shrink fit sleeve will usually stay put. If you are worried about that , fix it in position with a "dutchman" pin or screw. You can also build up the bore ID with brush plating.

We do not use "mild steel", it's 4140HT. Reasonably machinable but quite rigid & stable. Have changed some to fine grained cast iron though.

Obviously I've been at this too long. I think everything should be be built to my rather tight tolerances, I design machine tools, not toasters.LOL Yes, there are DIY short cuts, I just can't do that. (stubborn old fart)

Dick Z

I understand your point of view but sometimes the machine has to run, it is then that you look towards alternative fixes.

When dealing with a DIY solution though there are many more constraints on money, machinery and capability. The question often becomes how can I make this work with what I have and not break the bank. For this builder a Loctite approach might be an economical solution if the bore is in fact oversized. Further such a fix may last a very long time depending of course on his intended usage. A word of warning is in order though a even Loctite has limitations on what it can fix. Like fits for bearings it is always a good idea to refer to Loctites recommended practices.

Note that I only bring this up as a way to save this spindle and only if needed. Generally you don't want to use Loctite on anything you may want to disassemble from time to time. The need for a press and heat becomes a bit of a headache. On the other hand the search for perfection in a home shop often leads to nothing getting done.

[handlewanker]

Hi all, I think of all the rework processes, Loctite is the most user friendly.......provided the over size is not much more than a few thou, that is, the type of fit you get when you put the cut on to do a finishing cut and the vibration of the lathe motor makes the crosslide move just that bit more, or the crosslide is sticky and doesn't quite move the amount you put on, or the boring bar tool bit is blunt and doesn't cut when you do that .002" finish cut and you add another .001" and it takes the lot off....BTDT....a real pain.

The other alternative, boring out bigger and sleeving is labour intensive and also prone to the same error being repeated, but on several more fronts the next time around.

The other method, plating a thin layer of chrome with a brush method is interesting, seen it done for gold plating ornaments and bathroom tap ware, so you should be able to get the bore resized with a couple of thou of chrome.

BTW, when I referred to "mild steel", it is the common steel stock, supplied in cold rolled or hot rolled S1214 etc.............4140 is a nickel chrome alloy steel, pretty tough and not very user friendly to the DIY'er, but when you want that extra bit of toughness it'll do the trick nicely.

Manufacturers calculate the amount needed to retain their products, so should be followed for best results, but the rule of thumb method will siffice in all cases where press and clearance fits are required.

For instance, a bore size of 75mm (3" diam) by R.O.T (.001" per inch of diam) would indicate that the bore should be .003" less than the bearing OD for a general press fit.....if the bore was made 75mm dead, it would still be a press fit, (size for size), but not so tight.......so if the manufacturer's chart says .002" smaller, then you are within your fit either way.....having the rule of thumb means you are using a tried and trusted old school method that has worked for yonks.

You have to use a smidgeon of intelligence here for the application, IE, if you were making the housing from a block of alluminium you would use half the rule of thumb method for your press fit allowance or you'd distort the housing big time.

Loctite is a last resort when the best you did didn't happen......Boo Hoo, aww gee etc, LOL.
Ian.

[RICHARD ZASTROW]

Twice in the last year we had to replace spindle bearings. Bearings failed because of collisions. I'm glad they were not held by Loctite.

Dick Z

[handlewanker]

Hi, I guess the machine was of a quality that did not need this "fix me up".

I once had to remove the spindle and bearings from a revolving lathe centre.....the bearings were a tight press fit in the centre body with no holes at the back to allow a punch to knock them out......almost destroyed the revolving centre trying to get them out, but in the end I drilled the back of the body in two places opposit one another and used a punch against the outer races.
Ian.

[christofer]

Here are the ballscrews I have bought for this project so far. I have two of each screw (four screws total). These will be for the Y and Z axes.

The top screw is a 25mm diameter NSK ground screw with a C5 rating. The lead is about 16mm. I really like these screws but the ends are not machined. I have a feeling it will be a big hassle to machine the ends, thread them, and machine nuts for them. I paid $85 each for these screws. I like these screws because they are not used and they are very large and heavy duty looking

For this reason, I bought the lower screws in the pictures. These are made by THK and are used. They are 20 mm in diameter and have a 4mm lead. They are already machined and came with the bearings. The bearings are made by NSK and seem to have preload. I paid $139 each for these screws. I like these because I do not have to do any machining and I can avoid a belt drive.

I will measure the blacklash in all of these screws soon to be sure I will not have any problems later on once the choice is made.

Are both of these screws suitable for a machine of my size? Will the 20mm ones be too springy? Someone named Zach on this forum built a similar machine with the cheap C7 16mm ebay screws and it seems to work fine.